Sampling Procedure
The sample sites for the Freedom House Social Cohesion Survey 2016-2017 were drawn from the Combatting Drivers of Xenophobic Violence programme that had been running since March 2016.
The programme was conducted in 16 sites in 6 provinces (Eastern Cape, Gauteng, Kwa-Zulu Natal, Limpopo, North West and Western Cape) in South Africa, and is guided by Freedom House's working model of what is likely to lead to an 'at-risk' community.
The initial survey design was to conduct the same survey instrument as a baseline and endline 12 months apart in the same 5 sites in order to:
a) Inform the intervention to be conducted by the local MXV implementing partner peace building teams, and
b) Evaluate whether the intervention had had an effect.
The sites were selected purposively by FH in consultation with SSA. The sites were chosen to encapsulate the various contexts in terms of settings, that is, types of settlements (rural, peri-urban and urban) and geographical spread across provinces.
After the completion of the first phase in mid-2016 the design was changed so that the second phase would be implemented in five new sites rather than returning to the same sites. This was because:
a) The timing of the intervention and the study: the intervention had already been active for a longer period, so the survey field work could not be taken as a baseline
b) The timeframe of measuring impact would have to be longer than 12 months
c) Phase 1 field work took place just before local government elections, which is a very particular period in which to ask community residents about political participation and perceptions, so findings would not be comparable with other periods of time.
The total sample of sites was therefore 10 out of the 16 possible intervention sites.
Within each broad intervention site, specific wards or sections of wards were purposively sampled by FHSA, as shown in Table 1, based on the following criteria:
• Areas where the intervention was taking place
• Areas representing a diversity of settlement types within the site, e.g. informal and formal housing areas; areas with predominantly Black or 'Coloured' residents; commercial farming areas, township areas and formal 'town' areas.
• Excluding extremely dangerous or hard to access sites, such as the Glebeland Hostel in Durban South.
